A new study has found that SNP chips do not reliably identify rare disease-linked variants, CNN reports.
Researchers from the University of Exeter College of Medicine and Health conducted a retrospective analysis of the sensitivity and specificity of genotyping by SNP chip versus sequencing of nearly 50,000 people from the UK Biobank and the Personal Genome Project. As they report in the BMJ, SNP chips performed well and had high sensitivity, specificity, positive predictive value, and negative predictive value for more than 108,500 common variants.
However, the more rare the variant tested, the worse the approach performed, the researchers found. For instance, for pathogenic variants in BRCA1 and BRCA2, they report SNP chips had a sensitivity of 34.6 percent, a specificity of 98.3 percent, a positive predictive value of 4.2 percent, and a negative predictive value of 99.9 percent, as compared to sequencing.
"SNPs are just the wrong tool for the job when it comes to rare variants," study co-author Caroline Wright, professor of genomic medicine at Exeter, tells CNN. "They're excellent for the common variants that are present in lots of people, but the rarer the variant is, the less likely they are to be able to correctly detect it."